This thesis aims to improve the assessment of dam safety through the formalization of an identified failure mode that combines the sliding and overturning failure mechanisms. The failure mode is relevant to dams with geometrical variations and large-scale asperities in the rock-concrete interface and represents the behavior observed in finite element analyses (FEA) and scale model tests. 

Dams are vital infrastructure, providing services such as water storage for hydropower and irrigation, flood control, and containment of industrial byproducts. However, storing large volumes of water carries the risk of potentially catastrophic consequences in the event of failure, making safety assessments essential throughout a dam’s service life. Regulatory rules and guidelines prescribe methods for performing such dam safety assessments, in which a dam’s safety against global failure modes such as sliding and overturning is commonly evaluated. However, these failure modes rely on assumptions that may not be consistent with the actual behavior of dams when large-scale asperities are present along the rock-concrete interface, which may cause the dam to exhibit a failure mode that combines both sliding and overturning. Furthermore, these regulatory rules and guidelines typically employ deterministic assessment approaches, providing general safety factors that must be calibrated for specific cases to achieve the desired level of safety. Given that dams are unique structures with inherent uncertainties, this approach is unlikely to produce consistent safety levels.

The failure mode presented in this thesis, referred to as combined sliding and overturning (CSO), addresses the limitations of the traditional failure modes and reflects the failure behavior observed in concrete dams with large-scale asperities along the rock-concrete interface. Having been identified in FEA of concrete buttress dams, the failure mode was formalized and an analytical formulation was developed, providing results almost identical to those of FEA. The assumptions behind the failure mode and analytical formulation were compared with twelve scale model tests on concrete buttress dams, which included a monitoring system to validate these assumptions. The tests confirmed the assumptions and also evaluated the influence of factors such as rock bolts, reinforcement, and rock joints on load capacity and behavior, which primarily affected the load capacity. 

By relying solely on equilibrium equations, like traditional failure modes, the analytical formulation provides a simple alternative to methods such as FEA and can be readily applied with reliability analysis to assess a dam’s safety while accounting for its unique uncertainties. This is exemplified by a study on the reliability and sensitivity analysis of concrete buttress dams, in which a population of buttress dam monoliths, varying in height, width, and other characteristics, is generated and assessed for sliding, overturning, and CSO under two load cases, involving either overtopping or ice load. As such, influential factors for a dam’s reliability, including the basic friction angle, the dilation angle, the monolith height, and the inclinations of the front plate and the large-scale asperities, were identified. By introducing original methods to assess a dam’s load capacity while highlighting which factors impact the reliability of concrete dams, this thesis contributes to improved concrete dam safety assessment.